Traditional Chinese Herbal Formula XuMingZhuSan Alleviates Neuronal Ferroptosis in Rats After Stroke by Activating the NRF2/GPX4/SLC7A11 Pathway

ABSTRACT

This paper reported the efficacy of Traditional Chinese herbal formula XuMingZhuSan (XMZS) in stroke. Middle cerebral artery occlusion (MCAO) rat model was constructed and treated by XMZS and ML385. Brain sections of rats underwent histologic staining methods to assess brain infarction, neuronal and mitochondrial damage, and reactive oxygen species (ROS) production. Oxygen–glucose deprivation/reperfusion (OGD/R)‐induced hippocampal neurons were treated with XMZS and ML385. Apoptosis, ROS, mitochondrial membrane potential, and lipid peroxidation in hippocampal neurons were monitored by TUNEL, DCFH‐DA, JC‐1, and C11‐Bodipy staining. Ferroptosis‐related indicators in brains and hippocampal neurons were assayed using commercial kits. NRF2/GPX4/SLC7A11 pathway activity within brains and hippocampal neurons was assessed by Western blotting. In MCAO rats, XMZS improved neurological function; reduced cerebral infarction volume; relieved neuronal damage in the cortex and hippocampal CA1 and CA3 regions, and mitochondrial damage within the brain; suppressed ROS in the cortex and hippocampal CA1 and CA3 regions; decreased Fe and MDA contents within the brain; increased SOD and GSH within the brain; and activated NRF2/GPX4/SLC7A11 pathway within the brain. These influences induced by XMZS on MCAO rats were eliminated by ML385. XMZS mitigated apoptosis of OGD/R‐induced hippocampal neurons; reduced LDH activity and ROS; declined Fe and MDA contents; raised SOD and GSH levels; elevated mitochondrial membrane potential; suppressed lipid peroxidation; and activated NRF2/GPX4/SLC7A11 pathway. ML385 reversed these effects of XMZS on the OGD/R‐induced hippocampal neurons. XMZS may activate NRF2/GPX4/SLC7A11 pathway to treat stroke by attenuating neuronal ferroptosis and is applicable in clinical management of stroke.

Traditional Chinese herbal formula XuMingZhuSan showed its efficacy in treating stroke. It was capable of attenuating neuronal damage and mitochondrial damage and reducing infarction volume in rats with stroke. Mechanically, XuMingZhuSan might treat stroke by relieving neuronal ferroptosis via activating the NRF2/GPX4/SLC7A11 pathway.

1. Introduction

Stroke triggered by cerebral artery occlusion leads to severe disability and mortality worldwide, with incidence rates continuing to rise among younger populations (Nehme and Li 2025). Diabetes mellitus, hyperlipidemia, and hypertension are major risk factors for stroke (Qin et al. 2022). Stroke induces irreversible neuronal damage within a very short period of time from the onset, thereby leading to motor dysfunction and cognitive dysfunction in patients (Sasaibe et al. 2025). Thrombolytic therapy, thrombectomy, and antithrombotic therapy are common clinical interventions for stroke (Qin et al. 2022). However, there is still a lack of efficient approaches and drugs for improving neurological function recovery post‐stroke. There is an urgent need to develop effective therapies to alleviate brain and nervous system damage after stroke.

Traditional Chinese medicine is advantageous in treating stroke, such as effectiveness, safety, and affordability (Fan et al. 2024; Zhai et al. ²⁰²³; Zheng et al. ²⁰²⁴). Traditional Chinese herbal formula XuMingZhuSan (XMZS) contains 13 types of herbs, including Ephedra sinica Stapf, Cinnamomum cassia Presl, Asarum heterotropoides F. Schmidt, Heracleum hemsleyanum Diels, Radix Puerariae, Radix Ginseng, Angelica sinensis, Ligusticum chuanxiong, PaeonialactifloraPall., Rehmannia glutinosa Libosch., Nepeta cataria , Polygala tenuifolia Willd., and Saposhnikovia divaricata. In traditional Chinese medicine, XMZS is utilized in treating stroke, cerebral hemorrhage and transient ischemic attack, which is beneficial to the recovery of limb function in patients. In the field of Chinese medicine, XMZS has been approved for treating stroke, but more evidence is still needed to promote its widespread application in treating stroke. Middle cerebral artery occlusion (MCAO) animal model is usually adopted in simulating stroke in vivo (Goodman et al. 2023). Thus, this study performed MCAO surgery on rats to mimic an in vivo animal model of stroke and investigated the efficacy of XMZS in treating MCAO rats.

Ferroptosis is non‐apoptotic programmed cell death with a major mechanism of iron‐dependent lethal lipid peroxidation accumulation and disturbances in the antioxidant system (Li et al. 2024). Ferroptosis is a key factor in stroke prognosis because ferroptosis in the brain can cause and exacerbate neuronal damage (Guo et al. 2023; Hu et al. ²⁰²⁴). Anti‐ferroptosis strategy is thus proposed to be a promising strategy for stroke treatment (Tuo and Lei 2024; Xu et al. ²⁰²³; Wang et al. ²⁰²³). The Nuclear factor‐erythroid 2‐related factor‐2 (NRF2)/Glutathione peroxidase 4 (GPX4)/Solute Carrier Family 7 Member 11 (SLC7A11) pathway has been revealed to be associated with ferroptosis. Nrf2 is a transcription factor with an antioxidant effect, and it regulates the expression of GPX4 and SLC7A11, key components of ferroptosis. Nrf2 is thus considered a critical target for alleviating ferroptosis (Cui et al. 2025). As reported, the activation of the NRF2/GPX4/SLC7A11 pathway can exert neuroprotection by inhibiting ferroptosis after cerebral ischemia/reperfusion injury (Kang et al. 2025). In our preliminary study, XMZS medicated serum was discovered to reduce Fe level while activating the NRF2/GPX4/SLC7A11 pathway in the oxygen–glucose deprivation/reperfusion (OGD/R)‐induced primary hippocampal neurons. Currently, reports concerning XMZS are extremely rare. This work firstly explored whether XMZS could treat stroke by modulating the NRF2/GPX4/SLC7A11 pathway‐mediated ferroptosis. This study will provide a reliable theoretical basis for the application and promotion of XMZS in treating stroke.

2. Materials and Methods

2.1. Preparation of XMZS Decoction

XMZS formula consisted of the following ingredients: Ephedra sinica Stapf (9 g), Cinnamomum cassia Presl (12 g), Asarum heterotropoides F. Schmidt (6 g), Heracleum hemsleyanum Diels (15 g), Radix Puerariae (30 g), Radix Ginseng (10 g), Angelica sinensis (15 g), Ligusticum chuanxiong (9 g), PaeonialactifloraPall. (15 g), Rehmannia glutinosa Libosch. (10 g), Nepeta cataria (10 g), Polygala tenuifolia Willd. (12 g), and Saposhnikovia divaricata (10 g).

The XMZS decoction was prepared as follows: XMZS formula with a weight of 100 g was pulverized, introduced into cold water (500 mL) and steeped for 30 min. Thereafter, it was boiled for 40 min. The XMZS decoction was then collected by filtering. The remaining dregs were subjected to an additional 40 min of boiling within water (300 mL). Once again, the XMZS decoction was harvested by filtering. The decoctions obtained from both times were blended, concentrated (by boiling) to a final concentration of 3.15 g/mL, and preserved at 4°C before subsequent analysis.

2.2. Animals

We obtained 120 8‐week‐old male Sprague–Dawley (SD) rats (280 ± 20 g) in Vital River Laboratory Animal Technology (Beijing, China) and raised them within the animal laboratory room under 22°C with a 12‐h light/dark cycle. Each rat was allowed to take standard rodent chow and water unrestrictedly. Before the experiment, rats were subjected to adaptive feeding for 7 days.

Animal study had been ratified by the Animal Welfare Ethics Committee of Beijing MDKN Biotechnology Co. LTD. (MDKN‐2024‐041). As the authors' affiliated institution did not hold an animal experimentation license nor possess animal housing facilities, the animal experiments for this study were conducted at a company (i.e., Beijing MDKN Biotechnology Co. LTD.) holding both an animal license and an animal ethics committee. All experimental procedures were performed by the authors and their team. Animal experiments were carried out strictly following the experimental animal care and use guidelines of the Beijing Animal Control Committee.

2.3. MCAO Operation

Rats were deeply anesthetized via inhaling 2% isoflurane. After routine skin disinfection, the common carotid, external carotid artery, and internal carotid artery were sequentially exposed. After making an incision within the external carotid artery, we placed a 0.28‐mm nylon cord in the internal carotid artery gently for blocking the middle cerebral artery. Following 60 min, this nylon cord was withdrawn slightly for restoring blood flow. Then, rat skin was sutured. Rats were placed on a 37°C blanket until awakening.

2.4. Animal Treatment

Among the 120 rats, 72 rats were randomly selected and grouped as Sham, MCAO, MCAO + XMZS‐L, MCAO + XMZS‐M, MCAO + XMZS‐H, and MCAO + NBP groups (n = 12/group). These 72 rats were used for experiment 1. The remaining 48 rats were used for experiment 2, and randomized as Sham, MCAO, MCAO + XMZS, and MCAO + XMZS + ML385 groups (n = 12/group). All rats were randomly assigned to groups.

Before animal experiment, power analysis was performed to determine the number of rats in each group. With α value of 0.05% and 80% power, 8–10 rats in each group were required. To enhance the statistical power of the results, 12 rats were used per group.

The clinical dosage of XMZS was 350 g/70 kg/day. Based on the body surface area conversion formula between humans and rats, the equivalent dosage of XMZS for rats was: 350 g/70 kg/day × 6.3 = 31.5 g/kg/day. Thus, the dose of 31.5 g/kg/day was set as a moderate dose. Correspondingly, a low dose of XMZS for rats was: 31.5 g/kg/day × 0.5 = 15.75 g/kg/day. A high dose of XMZS for rats was: 31.5 g/kg/day × 2 = 63 g/kg/day. For each rat, 2 mL/100 g was administered. For rats treated by moderate and lower doses, XMZS decoction was diluted once and twice respectively before administration.

For experiment 1, treatments of rats were as follows: rats of Sham group received the surgery similar to MCAO, but without the blockage of the middle cerebral artery; those of MCAO group underwent MCAO surgery; those in MCAO + XMZS‐L, MCAO + XMZS‐M, and MCAO + XMZS‐H groups were firstly subjected to MCAO surgery and then administered with XMZS by gavage for seven consecutive days (15.75, 31.5, and 63 g/kg for the MCAO + XMZS‐L group, MCAO + XMZS‐M group, and MCAO + XMZS‐H group, respectively; once daily); for rats of the MCAO + NBP group (served as the positive control group), they were firstly subjected to MCAO surgery and then treated with Butylphthalide (NBP) (HY‐B0647, MCE, New Jersey, USA) by intraperitoneal injection (80 mg/kg; once daily).

For experiment 2, rats were treated as follows: rats of Sham group underwent the surgery similar to MCAO, but without the blockage of the middle cerebral artery; those of the MCAO group underwent MCAO surgery; the treatment of rats from the MCAO + XMZS group was identical to the MCAO + XMZS‐H group; those in the MCAO + XMZS + ML385 group firstly experienced MCAO surgery, and then were subjected to administration of XMZS (63 g/kg, once daily) and intraperitoneal injection of ML385 (Nrf2 inhibitor, 30 mg/kg, once daily) (Chen et al. 2023) for seven consecutive days.

After 7 days, rats were euthanized after being deeply anesthetized with 5% isoflurane. The whole brains were rapidly obtained and preserved at −80°C. Among the 12 rats per group, three rats were used for Triphenyltetrazolium chloride (TTC) staining, six for the detection of ferroptosis‐related indicators, and the remaining three for histological staining and Western blotting. No rat deaths occurred throughout the entire experimental period.

2.5. Neurological Function Assessment

After treatment for 7 days, the modified Neurological Stroke Scale (mNSS) and the modified Bederson Scale were utilized to evaluate rat neurological function. The mNSS scores ranged from 0 to 18, while Bederson scores ranged from 0 to 5. Higher mNSS and Bederson scores represented poorer neurological function. The detailed mNSS score criteria and Bederson score criteria could be found in previous reports (Leiss et al. 2022).

2.6. TTC Staining

The whole brains were prepared in 1‐mm sections, which were later immersed into 2% TTC solution (Biolab Technology, Beijing, China) to incubate 10 min at 37°C. The non‐infarction region showed red color, whereas the infarction region displayed white color. The infarction region volume was examined by Image Pro‐Plus 6.0 software (Media Cybernetics, Bethesda, MD, USA).

2.7. Hematoxylin–Eosin (H&E) Staining

After fixed with 4% paraformaldehyde (Biolab Technology, Beijing, China), brains of rats were paraffin‐embedded and prepared in 4‐μm sections. Dewaxing and rehydration were sequentially performed on the sections. After 5 min of hematoxylin staining (Biolab Technology, Beijing, China) and 1 min of eosin staining (Biolab Technology, Beijing, China) under ambient temperature, dehydration and transparency of the sections were implemented. The sections were sealed in neutral resin and observed under a light microscope (Olympus, Tokyo, Japan). Neuronal damage in the cortex, CA1, and CA3 regions was observed.

2.8. Immunofluorescence Staining

The brain sections (4 μm thickness) were blocked in 5% bovine serum albumin (Biolab Technology, Beijing, China) for 30 min, followed by 0.1% Triton X‐100 treatment (Biolab Technology, Beijing, China) for 15 min. For Neuronal nuclear antigen (NeuN) immunofluorescence staining, rabbit anti‐NeuN primary antibody (1:100, ab239347, Abcam, Cambridge, UK) was added to incubate the brain sections overnight under 4°C. Then Alexa Fluor 488‐conjugated goat anti‐rabbit secondary antibody (1:200, ab150077, Abcam, Cambridge, UK) was used to incubate the brain sections for 2 h under ambient temperature. To detect neuronal 4‐Hydroxynonenal (4‐HNE) expression, rabbit anti‐NeuN (1:100, ab239347, Abcam, Cambridge, UK) and mouse anti‐4‐HNE (MAB6115, AmyJet Scientific, Wuhan, China) primary antibodies were used to treat the brain sections. Alexa Fluor 555‐conjugated goat anti‐rabbit (ab150078, Abcam, Cambridge, UK) and 488‐conjugated goat anti‐mouse (SY0683, Biolab Technology, Beijing, China) secondary antibodies (1:200) were then employed to incubate the brain sections for 2 h under ambient temperature. Brain sections were stained with 4′, 6‐diamidino‐2‐phenylindole (DAPI) solution (Biolab Technology, Beijing, China) for 5 min. After drying and sealing, the brain sections were observed under a fluorescence microscope (Olympus, Tokyo, Japan).

2.9. Dihydroethidium (DHE) Staining

DHE staining of brain sections was employed to evaluate reactive oxygen species (ROS) level in the cortex and hippocampal CA1 and CA3 regions. Brain sections (4 μm thickness) were probed with dihydroethidium solution (MedChemExpress, New Jersey, USA) for 30 min at 37°C in darkness. DAPI was added onto the brain sections for nuclear staining. After being sealed, brain sections were monitored with the fluorescence microscope (Olympus, Tokyo, Japan). DHE staining in the cortex and hippocampal CA1 and CA3 regions was observed to analyze ROS level.

2.10. Transmission Electron Microscopy (TEM)

Ultrathin brain sections (60‐nm) were fixed in 4% paraformaldehyde and then underwent lead citrate and uranyl acetate staining. Mitochondrial structure was observed under TEM (Hitachi, Tokyo, Japan). The ruptured mitochondria percentage and mitochondrial length were analyzed to appraise mitochondrial damage.

2.11. Liquid Chromatography‐Mass Spectrometry (LC–MS)

A total of 12 male SD rats (purchased from Vital River Laboratory Animal Technology, Beijing, China) were randomly divided into Control group (n = 6) and Drug group (n = 6). Rats of the Drug group were administered with XMZS (63 g/kg, once daily) for 7 consecutive days. Those of the Control group were without any treatment during the 7‐day period. Blood samples of rats were harvested post 1 h of the last XMZS treatment. Following centrifugation (3000 rpm, 4°C, 20 min), serum samples were gathered.

To analyze the serum active components of XMZS, serum samples of rats in Control and Drug groups, as well as the filtered XMZS decoction, were subjected to LC–MS analysis. The procedure of LC–MS analysis was as follows: a mass spectrometer (AB Sciex, Foster City, CA, USA) equipped with an electrospray ionization source was coupled with a high performance liquid chromatography system (Shimadzu, Kyoto, Japan) were used. Serum samples (10 μL) and drug samples (10 μL) were separated by using a XBridge C18 column (4.6 × 100 mm, 3.5 μm). Solvent A was water with 0.1% formic acid, while Solvent B was acetonitrile with 0.1% formic acid. Samples were subjected to gradient elution for 30 min at a constant column temperature of 35°C. In the following, the system was analyzed in both positive and negative ion modes, with a mass range scan of 100–1500 m/z. Data processing was implemented by employing Progenesis QI (Version 2.4) software. This method was referred to a previous study (Chai et al. 2025). The chromatography column mobile phase gradient parameters were shown in Table 1. The re‐equilibration step time was 2 min and total run time was 10 min. MS source settings were as follows: ion source: electrospray ionization source; capillary voltage: 3.0 kV for positive ions and 2.5 kV for negative ions; gas flows: 150 L/h for cone hole gas flow rate and 1000 L/h for desolventizing gas flow rate; source temperature: 120°C. Mass resolution and calibration/lock‐mass were as follows: real‐time calibration was performed using leucine enkephalin (m/z: 556.2771 for positive ion, and 554.2615 for negative ion); calibration was conducted using sodium formate; the resolution was 6 W. Acquisition mode was MSe. Collision energy was 20–80 V.

TABLE 1.

Chromatography column mobile phase gradient parameters.

Time (min)	Flow rate (mL/min)	Solvent A (%)	Solvent B (%)
Initial	0.4	95	5
0.5	0.4	95	5
1.0	0.4	80	20
4.0	0.4	40	60
6.0	0.4	1	99
8.0	0.4	1	99
8.1	0.4	95	5
10.0	0.4	95	5

2.12. Preparation of XMZS Medicated Serum

A total of 20 male SD rats (weights, 220 ± 20 g) were acquired from Vital River Laboratory Animal Technology and given 7 consecutive days of XMZS (63 g/kg, once daily) via gavage. After 1 h of the last XMZS administration, blood from all rats was gathered from the abdominal aorta and left for a 3‐h duration under ambient temperature. Serum was harvested following 20 min of blood centrifugation (3000 rpm, 4°C). Serum samples were subjected to an inactivation process at 56°C for 30 min and then mixed together. The mixed serum sample was preserved at −80°C. Before using, the inactivated serum sample was diluted to different volume percentages.

2.13. Isolation of Primary Hippocampal Neurons From Newborn Rats

Newborn rats were obtained from Vital River Laboratory Animal Technology (Beijing, China). Hippocampal tissues were isolated from the newborn rats, and then cut into small pieces. The small pieces were incubated with 0.125% trypsin (Biolab Technology, Beijing, China) for 15 min at 37°C. Dissociated neurons were gathered by centrifugation, and then cultured in Neurobasal medium (Gibco, New York, USA) at 37°C with 5% CO₂. The medium was refreshed at a 2‐day interval.

2.14. Identification of Primary Hippocampal Neurons

Immunofluorescence staining of neuronal markers, including NeuN and Mitogen‐activated protein (MAP), was employed for the identification of the isolated primary hippocampal neurons. After culture for 5 days, primary hippocampal neurons were fixed within 4% paraformaldehyde for 10 min. Rabbit anti‐NeuN (ab239347, Abcam, Cambridge, UK) and mouse anti‐MAP (MAB11155, AmyJet Scientific, Wuhan, China) primary antibodies (1:100), as well as Alexa Fluor 488‐conjugated goat anti‐rabbit (ab150077) and 555‐conjugated goat anti‐mouse (ab150114) secondary antibodies (1:200, Abcam, Cambridge, UK) were used to treat primary hippocampal neurons. DAPI was utilized for nuclear staining. Positive expression of NeuN and MAP was observed under the fluorescence microscope (Olympus, Tokyo, Japan), indicating successful isolation of primary hippocampal neurons.

2.15. OGD/R Induction

OGD/R induction was carried out on primary hippocampal neurons to simulate the in vitro stroke model. After growing primary hippocampal neurons in glucose‐ and serum‐free medium under 5% CO₂, 95% N₂, and 37°C conditions for 1 h, they were grown within medium containing glucose and 10% fetal bovine serum and kept under a 5% CO₂, 95% air, and 37°C conditions for 24 h.

2.16. Primary Hippocampal Neuron Treatment

After OGD/R induction, primary hippocampal neurons were cultured for 48 h in normal medium containing XMZS medicated serum (with volume percentages of 5%, 10%, 15%, 20%, 30%, 40%, and 50%) at 37°C with 5% CO₂. XMZS medicated serum (10% volume percentage) combined with ML385 (Nrf2 inhibitor, 1 μM) (Chen et al. 2023) were used at the same time to treat the OGD/R‐induced primary hippocampal neurons at 37°C with 5% CO₂ for 48 h. Primary hippocampal neurons cultivated under normal conditions were utilized as control.

2.17. Cell Counting Kit‐8 (CCK‐8) Assay

Primary hippocampal neurons were seeded into 96‐well plates with 5 × 10³ cells per well, and subsequently incubated with CCK‐8 reagent (10 μL, Biolab Technology, Beijing, China) for 2 h at 37°C. Absorbance in every well was monitored with the microplate reader (Biotek, Winooski, VT, USA), which was utilized to evaluate cell viability.

2.18. Terminal Deoxynucleotidyl Transferase dUTP Nick‐End Labeling (TUNEL) Assay

Primary hippocampal neurons were fixed with 4% paraformaldehyde for 10 min and then treated with 0.1% Triton X‐100 permeabilization for 10 min. After 1 h of staining by TUNEL working solution (Zeye Biotechnology, Shanghai, China) at 37°C and nuclear staining by DAPI, primary hippocampal neurons were monitored under the fluorescence microscope (Olympus, Tokyo, Japan). Apoptosis rate was determined by the Image Pro‐Plus 6.0 software.

2.19. Ferroptosis‐Related Indicators and Lactate Dehydrogenase (LDH) Activity

Brain tissues from rats and primary hippocampal neurons were collected, followed by 30 min of lysis on ice. Lysis mixture samples underwent centrifugation (12,000 rpm, 4°C, 15 min) to gather the supernatant. Ferroptosis‐related indicators within supernatants were analyzed using commercial kits (Jingkang Bioengineering, Shanghai, China), including Fe, malondialdehyde (MDA), superoxide dismutase (SOD), and glutathione (GSH). A LDH activity assay kit (Kanglang Biotechnology, Shanghai, China) was employed for testing LDH activity within the supernatant.

2.20. Dichlorodihydrofluorescein Diacetate (DCFH‐DA) Staining

To explore ROS level, primary hippocampal neurons were incubated with DCFH‐DA probe (Biolab Technology, Beijing, China) for 20 min at 37°C away from light. Under the fluorescence microscope (Olympus, Tokyo, Japan), ROS fluorescence densities were monitored and quantified with Image Pro‐Plus 6.0 software.

2.21. Western Blotting

Rat brain tissues and primary hippocampal neurons were collected. Radio‐immuno precipitation assay lysis buffer (Biolab Technology, Beijing, China) was adopted to treat tissues and neurons on ice for 30 min, in order to extract total proteins. After 15 min of centrifugation (12,000 rpm, 4°C), supernatants containing total proteins were harvested. Total protein levels in the supernatants were monitored with the BCA protein assay kit (Biolab Technology, Beijing, China). Later, protein aliquots (50 μg) were separated through 10% sodium dodecyl sulfate‐polyacrylamide gel electrophoresis prior to transfer to polyvinylidene fluoride (PVDF) membranes. Following 1 h of blockage within 5% defatted milk under ambient temperature, PVDF membranes were incubated using primary antibodies for 12 h at 4°C. Primary antibodies included: anti‐Solute Carrier Family 7 Member 11 (SLC7A11) (1:1000, DF12509, Affinity Biosciences, USA); anti‐Nuclear factor‐erythroid 2‐related factor‐2 (NRF2) (16396‐1‐AP), anti‐Glutathione peroxidase 4 (GPX4) (ab125066), and anti‐β‐catenin (66009‐1‐Ig) (1:1000, Proteintech, Chicago, IL, USA). Thereafter, corresponding secondary antibodies (1:5000, Abcam, Cambridge, UK) were dropped and reacted under ambient temperature for 1 h. Protein blots were developed by treating PVDF membranes with enhanced chemiluminescent solution (Biolab Technology, Beijing, China). The Image Pro‐Plus 6.0 software was employed for quantifying protein blots. β‐catenin was regarded as the reference.

2.22. JC‐1 Staining

Changes of mitochondrial membrane potential in primary hippocampal neurons were monitored by JC‐1 staining. Primary hippocampal neurons were stained with JC‐1 working solution staining (Biolab Technology, Beijing, China) for 20 min at 37°C in a dark environment, and then were observed under the fluorescence microscope (Olympus, Tokyo, Japan). The JC‐1 polymers (red fluorescence)/JC‐1 monomers (green fluorescence) ratio was determined for appraising mitochondrial membrane potential. The Image Pro‐Plus 6.0 software was adopted for analyzing fluorescence intensity.

2.23. C11‐Bodipy Staining

Primary hippocampal neurons underwent 30 min of incubation using C11‐Bodipy 581/591 probe (MedChemExpress, New Jersey, USA) at 37°C away from light. The C11‐Bodipy 581/591 probe shifted from red fluorescence to green fluorescence in the presence of lipid peroxidation, as monitored with a fluorescence microscope (Olympus, Tokyo, Japan). The ratio of green fluorescence/red fluorescence, qualified with Image Pro‐Plus 6.0 software, was utilized for evaluating lipid peroxidation.

2.24. Statistical Analysis

All experiments were performed in biological replicates. Measurement results were acquired from three separate assays and represented by mean ± standard deviation. Distinct groups (at least three groups) were compared by one‐way ANOVA with post hoc Tukey's test. Statistical analyses were accomplished with GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA). p < 0.05 stood for significant differences.

3. Results

3.1. Analysis of Serum Active Components of XMZS

Using LC–MS, serum active components of XMZS were assayed. The base peak chromatograms of control serum, drug serum, and drug in positive and negative ion modes were displayed (Figures S1 and S2). A total of 25 serum active components of XMZS were screened in rats, including Inden, Tyrosine, Phenylalanine, etc. (Figure S3).

3.2. XMZS Reduced Infarction Volume and Relieved Neuronal Damage of MCAO Rats

Animal experimental flow was shown in Figure 1A. After a 7‐day treatment, MCAO rats possessed distinctly higher mNSS and Bederson scores, comparatively (MCAO vs. Sham groups) (p < 0.001). This result indicated the successful construction of MCAO models. Compared with MCAO rats without any treatment, MCAO rats treated by low and moderate doses of XMZS showed lower Bederson score (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups) (p < 0.05). However, low and moderate doses of XMZS had minimal effect on mNSS score in MCAO rats (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups). Notably, high dose of XMZS treatment decreased mNSS and Bederson scores in MCAO rats (MCAO + XMZS‐H vs. MCAO groups) (p < 0.01). As a positive control, NBP treatment yielded results similar to those observed with high dose of XMZS administration, significantly reducing mNSS and Bederson scores in MCAO rats (MCAO + NBP vs. MCAO groups) (p < 0.01) (Figure 1B,C). By TTC staining, large infarction volume was found in MCAO rats (MCAO vs. Sham groups) (p < 0.0001), suggesting a successful construction of MCAO models. Low and moderate doses of XMZS treatment had no obvious influence on infarction volume in MCAO rats (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups). High dose of XMZS treatment prominently reduced infarction volume in MCAO rats (MCAO + XMZS‐H vs. MCAO groups) (p < 0.01), with efficacy comparable to NBP treatment (MCAO + NBP vs. MCAO groups) (p < 0.01) (Figure 1D,E). From H&E images, rats of the Sham group showed normal neurons in the cortex, CA1, and CA3 regions, but severe neuronal damage occurred in rats of the MCAO group. XMZS treatment provided obvious relief on neuronal damage in the cortex, CA1, and CA3 regions of MCAO rats, especially at high dose (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups; MCAO + XMZS‐H vs. MCAO groups). NBP showed similar effect to high dose of XMZS, obviously alleviating neuronal damage in the cortex, CA1, and CA3 regions of MCAO rats (MCAO + NBP vs. MCAO groups) (Figure 1F). Herein, these results showed the effectiveness of XMZS on decreasing infarction volume while attenuating neuronal damage of MCAO rats.

FIGURE 1.

XMZS reduced infarction volume and mitigated neuronal damage in MCAO rats. (A) Experimental flow of animal study. (B, C) XMZS treatment declined mNSS and Bederson scores in MCAO rats. (D, E) TTC staining indicated that XMZS treatment diminished infarction volume in MCAO rats. (F) H&E staining showed that XMZS treatment mitigated neuronal damage in the cortex, CA1, and CA3 regions of MCAO rats. *p < 0.05. **p < 0.01. ***p < 0.001. ****p < 0.0001. The symbol “ns” represented no statistically significant difference between groups.

3.3. XMZS Attenuated Ferroptosis in Brain Tissues of MCAO Rats

To appraise ferroptosis, Fe, MDA, SOD, and GSH contents within rat brain tissues were examined. MCAO rats exhibited elevated Fe and MDA contents whereas decreased SOD and GSH contents within brain tissues, comparatively (MCAO vs. Sham groups) (p < 0.001). Such reduced Fe and MDA contents, along with elevated SOD and GSH contents, could be detected from brain tissues of MCAO rats treated by moderate and high doses XMZS (MCAO + XMZS‐M vs. MCAO groups; MCAO + XMZS‐H vs. MCAO groups) (p < 0.05, p < 0.01, p < 0.001). However, the effect of low dose of XMZS treatment on Fe, MDA, SOD, and GSH contents within brain tissues of MCAO rats was not statistically significant (MCAO + XMZS‐L vs. MCAO groups). Similar to high dose of XMZS treatment, NBP treatment markedly decreased Fe and MDA contents while increasing SOD and GSH contents within brain tissues of MCAO rats (p < 0.01, p < 0.001) (Figure 2A–D). Immunofluorescence (Figure 2E) and DHE staining (Figure 2F) showed the attenuated NeuN staining but the intensified ROS staining within cortex, CA1, and CA3 regions in MCAO rats (MCAO vs. Sham groups). XMZS treatment, especially at high dose, obviously enhanced NeuN staining but attenuated ROS staining within cortex, CA1, and CA3 regions in MCAO rats (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups; MCAO + XMZS‐H vs. MCAO groups). NBP treatment exhibited similar effect to high dose of XMZS treatment, which obviously intensified NeuN staining and attenuated ROS staining within cortex, CA1, and CA3 regions in MCAO rats (MCAO + NBP vs. MCAO groups). Hence, XMZS relieved ferroptosis within brain tissues in MCAO rats.

FIGURE 2.

XMZS suppressed ferroptosis in brain tissues of MCAO rats. (A–D) XMZS relieved ferroptosis in brain tissues of MCAO rats, as it reduced Fe and MDA contents and elevated SOD and GSH contents in brain tissues of MCAO rats. (E) By immunofluorescence staining, XMZS enhanced NeuN immunofluorescence staining in the cortex, CA1, and CA3 regions of MCAO rats. (F) By DHE staining, XMZS reduced ROS staining in the cortex, CA1, and CA3 regions of MCAO rats. *p < 0.05. **p < 0.01. ***p < 0.001.

3.4. XMZS Activated NRF2/GPX4/SLC7A11 Pathway and Relieved Mitochondrial Damage Within Brain Tissues in MCAO Rats

The ferroptosis‐related NRF2/GPX4/SLC7A11 pathway activity was examined by Western blotting. NRF2, GPX4 and SLC7A11 proteins were prominently down‐regulated within brain tissues in MCAO rats (MCAO vs. Sham groups) (p < 0.001). Low dose of XMZS administration of MCAO rats remarkably increased NRF2 and GPX4 proteins within brain tissues (p < 0.05, p < 0.001), whereas it did not obviously affect the expression of SLC7A11 protein (MCAO + XMZS‐L vs. MCAO groups). The expression of NRF2, GPX4 and SLC7A11 proteins within brain tissues in MCAO rats was elevated after moderate and high doses of XMZS administration (MCAO + XMZS‐M vs. MCAO groups; MCAO + XMZS‐H vs. MCAO groups) (p < 0.05, p < 0.01, p < 0.001). Additionally, NBP treatment significantly up‐regulated the expression of NRF2, GPX4 and SLC7A11 proteins within brain tissues in MCAO rats (MCAO + NBP vs. MCAO groups) (p < 0.001), which was similar to high dose of XMZS treatment (Figure 3A–D). According to 4‐HNE/NeuN immunofluorescence staining, the increased 4‐HNE+/NeuN+ positive neurons within brain tissues in MCAO rats was observed (MCAO vs. Sham groups) (p < 0.001), while XMZS administration reduced it (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups; MCAO + XMZS‐H vs. MCAO groups) (p < 0.05, p < 0.01). NBP treatment distinctly decreased 4‐HNE+/NeuN+ positive neurons within brain tissues in MCAO rats (MCAO + NBP vs. MCAO groups) (p < 0.01), demonstrating an effect comparable to high dose of XMZS administration (Figure 3E,F). Based on TEM images, a higher ruptured mitochondria percentage and shorter mitochondrial length was found in MCAO rats (MCAO vs. Sham groups) (p < 0.001). Low and moderate XMZS administration presented negligible effects on ruptured mitochondria percentage and mitochondrial length in MCAO rats (MCAO + XMZS‐L vs. MCAO groups; MCAO + XMZS‐M vs. MCAO groups). High dose of XMZS was effective in reducing ruptured mitochondria percentage and elevating mitochondrial length (MCAO + XMZS‐H vs. MCAO groups) (p < 0.05, p < 0.01). This effect of high dose of XMZS was similar to that of NBP treatment, as NBP administration significantly reduced ruptured mitochondria percentage and extended mitochondrial length (MCAO + NBP vs. MCAO groups) (p < 0.05, p < 0.01) (Figure 3G–I). Thus, XMZS activated NRF2/GPX4/SLC7A11 pathway and mitigated mitochondrial damage within brain tissues in MCAO rats.

FIGURE 3.

XMZS activated NRF2/GPX4/SLC7A11 pathway and mitigated mitochondrial damage in brain tissues of MCAO rats. (A–D) By Western blotting, the NRF2/GPX4/SLC7A11 pathway was inactivated in brain tissues of MCAO rats, but was activated after XMZS treatment. (E‐F) 4‐HNE/NeuN immunofluorescence staining displayed the increased neuronal 4‐HNE expression in brain tissues of MCAO rats, but XMZS decreased its expression. (G–I) TEM images displayed that XMZS relieved mitochondrial damage in brain tissues of MCAO rats, as it reduced ruptured mitochondria percentage and increased mitochondrial length. The capital letter “M” stood for mitochondria. *p < 0.05. **p < 0.01. ***p < 0.001. ****p < 0.0001. The symbol “ns” represented no statistically significant difference between groups.

3.5. XMZS Medicated Serum Might Alleviate the OGD/R‐Induced Ferroptosis in Primary Hippocampal Neuron via Activating NRF2/GPX4/SLC7A11 Pathway

The isolated primary hippocampal neurons in newborn rats were identified through immunofluorescence staining because they could express the neuronal markers NeuN and MAP (Figure 4A). Primary hippocampal neurons were induced by OGD/R to mimic an in vitro stroke model and then treated with different volume percentages of XMZS medicated serum. OGD/R apparently decreased the primary hippocampal neuron viability (OGD/R vs. Control groups) (p < 0.001). XMZS medicated serum with volume percentages of 5%, 10%, 15%, and 20% could improve OGD/R‐induced primary hippocampal neuron viability significantly (p < 0.05, p < 0.01, p < 0.001). Notably, XMZS medicated serum of 10% volume percentage had the most pronounced improving effect on OGD/R‐induced primary hippocampal neuron viability (Figure 4B). Thus, in the subsequent in vitro mechanism study, XMZS medicated serum with 10% volume percentage was selected to treat the primary hippocampal neurons under OGD/R induction.

FIGURE 4.

XMZS medicated serum might attenuate the OGD/R‐induced ferroptosis in primary hippocampal neurons by activating NRF2/GPX4/SLC7A11 pathway. (A) The isolated primary hippocampal neurons from newborn rats were identified by immunofluorescence staining, as they were capable of expressing the neuronal markers NeuN and MAP. (B) CCK‐8 assay suggested that XMZS medicated serum with 10% volume percentage was the optimal dose to treat the OGD/R‐induced primary hippocampal neurons. (C) XMZS medicated serum inhibited LDH activity in the OGD/R‐induced primary hippocampal neurons, but was reversed by ML385. (D, E) By TUNEL staining, XMZS medicated serum relieved apoptosis in the OGD/R‐induced primary hippocampal neurons, which was abrogated by ML385. (F–I) XMZS medicated serum mitigated ferroptosis in the OGD/R‐induced primary hippocampal neurons (as shown by the decreased Fe and MDA contents and the increased SOD and GSH contents), but was counteracted by ML385. (J‐K) According to DCFH‐DA staining, XMZS medicated serum suppressed ROS fluorescence intensity in the OGD/R‐induced primary hippocampal neurons, but ML385 abolished this inhibitory effect. *p < 0.05. **p < 0.01. ***p < 0.001.

XMZS medicated serum (10% volume percentage) combined with NRF2 inhibitor ML385 was employed simultaneously to treat OGD/R‐induced primary hippocampal neurons. OGD/R induction led to increased LDH activity and apoptosis rate in primary hippocampal neurons (OGD/R vs. Control groups) (p < 0.001). XMZS medicated serum treatment decreased LDH activity and apoptosis rate of OGD/R‐induced primary hippocampal neurons (OGD/R + MS‐10% vs. OGD/R groups) (p < 0.001). However, these influences induced by XMZS medicated serum were abolished by ML385 (OGD/R + MS‐10% + ML385 vs. OGD/R + MS‐10% groups) (p < 0.05, p < 0.001) (Figure 4C–E).

OGD/R induction caused the elevated Fe and MDA contents but decreased SOD and GSH concentrations within primary hippocampal neurons (OGD/R vs. Control groups) (p < 0.001). In contrast, XMZS medicated serum diminished Fe and MDA contents, and enhanced SOD and GSH activities within OGD/R‐induced primary hippocampal neurons (OGD/R + MS‐10% vs. OGD/R groups) (p < 0.01, p < 0.001). ML385 treatment eliminated this effect of XMZS medicated serum on these ferroptosis‐related indicators within the OGD/R‐induced primary hippocampal neurons (OGD/R + MS‐10% + ML385 vs. OGD/R + MS‐10% groups) (p < 0.05, p < 0.01) (Figure 4F–I).

By DCFH‐DA staining, the increased ROS fluorescence intensity was detected within OGD/R‐induced primary hippocampal neurons (OGD/R vs. Control groups) (p < 0.001). XMZS medicated serum significantly reduced ROS fluorescence intensities within OGD/R‐induced primary hippocampal neurons (OGD/R + MS‐10% vs. OGD/R groups) (p < 0.001). However, the suppression of XMZS medicated serum against ROS levels within OGD/R‐induced primary hippocampal neurons could be reversed after ML385 treatment (OGD/R + MS‐10% + ML385 vs. OGD/R + MS‐10% groups) (p < 0.01) (Figure 4J,K). Therefore, XMZS medicated serum might activate NRF2/GPX4/SLC7A11 pathway to attenuate the OGD/R‐induced primary hippocampal neuronal ferroptosis.

3.6. XMZS Medicated Serum Might Attenuate the OGD/R‐Induced Mitochondrial Damage and Lipid Peroxidation of Primary Hippocampal Neurons by Activating NRF2/GPX4/SLC7A11 Pathway

According to Western blotting, OGD/R induction apparently down‐regulated NRF2, GPX4, and SLC7A11 protein levels inside primary hippocampal neurons (OGD/R vs. Control groups) (p < 0.001). XMZS medicated serum distinctly up‐regulated NRF2, GPX4, and SLC7A11 protein levels inside OGD/R‐induced primary hippocampal neurons (OGD/R + MS‐10% vs. OGD/R groups) (p < 0.001). ML385 eliminated this effect of XMZS medicated serum (OGD/R + MS‐10% + ML385 vs. OGD/R + MS‐10% groups) (p < 0.001) (Figure 5A–D).

FIGURE 5.

XMZS medicated serum might relieve the OGD/R‐induced mitochondrial damage and lipid peroxidation in primary hippocampal neurons by activating the NRF2/GPX4/SLC7A11 pathway. (A–D) According to Western blotting, XMZS medicated serum activated NRF2/GPX4/SLC7A11 pathway in the OGD/R‐induced primary hippocampal neurons, whereas ML385 eliminated this effect of XMZS medicated serum. (E, F) By JC‐1 staining, XMZS medicated serum increased mitochondrial membrane potential in the OGD/R‐induced primary hippocampal neurons, but ML385 abrogated this influence. (G, H) C11‐Bodipy staining illustrated the suppression of XMZS medicated serum on lipid peroxidation in the OGD/R‐induced primary hippocampal neurons, which was reversed by ML385. ***p < 0.001.

JC‐1 staining (Figure 5E,F) and C11‐Bodipy staining (Figure 5G,H) were employed to detect mitochondrial membrane potential changes and lipid peroxidation extent. After OGD/R stimulation, the declined mitochondrial membrane potential but enhanced lipid peroxidation was discovered in primary hippocampal neurons (OGD/R vs. Control groups) (p < 0.001). XMZS medicated serum elevated mitochondrial membrane potential and reduced lipid peroxidation of OGD/R‐induced primary hippocampal neurons (OGD/R + MS‐10% vs. OGD/R groups) (p < 0.001). These influences induced by XMZS medicated serum against mitochondrial membrane potential and lipid peroxidation of OGD/R‐induced primary hippocampal neurons were abolished by ML385 (OGD/R + MS‐10% + ML385 vs. OGD/R + MS‐10% groups) (p < 0.001). Hence, XMZS medicated serum could activate NRF2/GPX4/SLC7A11 pathway to attenuate OGD/R‐induced mitochondrial damage and lipid peroxidation of primary hippocampal neurons.

3.7. XMZS Might Relieve Neuronal Damage and Ferroptosis Within Brain Tissues in MCAO Rats Through Activating NRF2/GPX4/SLC7A11 Pathway

In vivo study was performed by administering MCAO rats with a high dose of XMZS and ML385. As shown in Figure 6A,B, the increased mNSS and Bederson scores in MCAO rats (MCAO vs. Sham groups) (p < 0.001) declined after XMZS administration (MCAO + XMZS vs. MCAO groups) (p < 0.05, p < 0.001). However, the downward impact of XMZS against mNSS and Bederson scores in MCAO rats was eliminated by ML385 (MCAO + XMZS + ML385 vs. MCAO + XMZS groups) (p < 0.05).

FIGURE 6.

XMZS might alleviate neuronal damage and ferroptosis within brain tissues in MCAO rats via activating NRF2/GPX4/SLC7A11 pathway. (A, B) XMZS administration reduced mNSS and Bederson scores in MCAO rats, which was reversed by ML385. (C–F) The suppression of XMZS on ferroptosis in brain tissues of MCAO rats was eliminated by ML385. (G) Based on immunofluorescence staining, XMZS intensified NeuN immunofluorescence staining in the cortex, CA1, and CA3 regions of MCAO rats, whereas ML385 reversed this effect of XMZS. (H) As shown by DHE staining, XMZS attenuated ROS staining in the cortex, CA1, and CA3 regions of MCAO rats, but this influence was counteracted by ML385. (I–K) According to TEM images, the mitigation effect of XMZS on mitochondrial damage in brain tissues of MCAO rats was abolished by ML385. The capital letter “M” stood for mitochondria. *p < 0.05. **p < 0.01. ***p < 0.001.

MCAO rats showed higher Fe and MDA contents but lower SOD and GSH activities within brain tissues (MCAO vs. Sham groups) (p < 0.01, p < 0.001). XMZS treatment resulted in a decrease in Fe and MDA contents, and an increase in SOD and GSH contents within brain tissues in MCAO rats (MCAO + XMZS vs. MCAO groups) (p < 0.01, p < 0.001). Compared with MCAO rats treated with XMZS (MCAO + XMZS group), those treated with XMZS and ML385 displayed elevated Fe and MDA contents but decreased SOD and GSH contents within brain tissues (MCAO + XMZS + ML385 vs. MCAO + XMZS groups) (p < 0.05, p < 0.01) (Figure 6C–F).

According to immunofluorescence (Figure 6G) and DHE staining (Figure 6H), MCAO rats exhibited the attenuated NeuN staining but the intensified ROS staining within cortex, CA1, and CA3 regions (MCAO vs. Sham groups). After being administered with XMZS, the enhanced NeuN staining along with the weakened ROS staining was detected within cortex, CA1, and CA3 regions in MCAO rats (MCAO + XMZS vs. MCAO groups). However, referred to MCAO + XMZS group, rats in MCAO + XMZS + ML385 group exhibited the attenuated NeuN staining and the intensified ROS staining in the cortex, CA1, and CA3 regions (MCAO + XMZS + ML385 vs. MCAO + XMZS groups) (Figure 6G,H).

Under TEM, shorter mitochondrial length along with higher ruptured mitochondria percentage occurred within brain tissue in MCAO rats (MCAO vs. Sham groups) (p < 0.001). By XMZS administration, mitochondrial length increased and ruptured mitochondria percentage declined within brain tissues in MCAO rats (MCAO + XMZS vs. MCAO groups) (p < 0.01). However, relative to MCAO + XMZS group, rats in MCAO + XMZS + ML385 group presented shorter mitochondrial length and higher ruptured mitochondria percentage in brain tissues (p < 0.05, p < 0.01) (Figure 6I–K). Thus, ML385 reversed the relief of XMZS on neuronal damage and ferroptosis within brain tissues in MCAO rats. This suggested that XMZS might activate NRF2/GPX4/SLC7A11 pathway to mitigate neuronal damage and ferroptosis within brain tissues in MCAO rats.

4. Discussion

This paper reported the therapeutic effects of XMZS, a Traditional Chinese herbal formula, on stroke. The evidence suggested that XMZS reduced cerebral infarction volume, improved neurological function and alleviated neuronal damage in MCAO rats. XMZS might activate NRF2/GPX4/SLC7A11 pathway to relieve neuronal damage in MCAO rats by suppressing ferroptosis. As we know, the efficacy induced by XMZS against stroke and related molecular mechanism was elucidated for the first time. Furthermore, this paper analyzed the serum active components of XMZS via LC–MS, which has advanced the clinical translational application value of XMZS research outcomes.

Ferroptosis is related to brain damage and neuronal injury after stroke, which is recognized as a key mechanism for a range of pathological responses after stroke (Liu, Yang, et al. ²⁰²⁴; Fan et al. ²⁰²³; Liu, Wang, et al. ²⁰²⁴). The mechanisms of ferroptosis can be summarized as an imbalance in the metabolism of iron, lipid peroxidation, and GSH (Tian et al. 2024). Ferroptosis exhibits the representative feature of iron‐dependent lipid peroxidation, which induces excessive MDA and ROS accumulation; however, MDA and ROS can react with proteins or DNA to produce toxic effects (Zhou et al. 2023). GSH belongs to the antioxidant system that inhibits progression of ferroptosis through blocking lipid peroxidation (Liu, Cui, et al. ²⁰²³). SOD is capable of removing oxygen radicals to mitigate cellular damage in ferroptosis (Zhang et al. 2022). As one of the major byproducts of lipid peroxidation, 4‐HNE can disrupt the structure and function of DNA and proteins by interacting with them (Liu, Pang, et al. ²⁰²³). In this study, XMZS was suggested to be effective in suppressing ferroptosis after stroke, as it reduced Fe, MDA, ROS, lipid peroxidation, 4‐HNE, and increased SOD and GSH of MCAO rats and the OGD/R‐induced primary hippocampal neurons.

Unlike other cell death patterns, the cellular morphology changes in ferroptosis are primarily featured by the decreased mitochondrial size, shrinking mitochondrial membrane density, reduced and disappeared mitochondrial cristae, and rupturing outer mitochondrial membrane (Tian et al. 2024). In this paper, XMZS treatment remarkably declined the ruptured mitochondria percentage and elevated mitochondrial length in brains of MCAO rats. One of the other hallmarks of the ferroptosis‐induced mitochondrial damage is the reduced mitochondrial membrane potential (Gao et al. 2019). Based on this article, mitochondrial membrane potential was significantly declined within the OGD/R‐induced primary hippocampal neurons, while XMZS treatment elevated it. Thus, XMZS was effective on alleviating ferroptosis‐induced mitochondrial damage after stroke.

Stroke is a life‐threatening disease in which mitochondrial damage and dysfunction represent a key pathological hallmark in the brain (Wang et al. 2025; You et al. ²⁰²⁵). Mitochondrial damage severely impairs neuronal survival and neurological recovery following stroke (She et al. 2025). Mitochondrial dysfunction causes disruption of the mitochondrial membrane potential and neuronal apoptosis (Guo et al. 2025). Besides, the dissipation of mitochondrial membrane potential induced by ferroptosis plays a pivotal role in the pathogenesis and progression of stroke (Gong et al. 2025). Recent reports have reviewed that some traditional Chinese medicines can mitigate stroke in animals by relieving mitochondrial damage and dysfunction via blocking ferroptosis (Huang et al. 2025; Ma et al. ²⁰²⁵). Currently, the role of XMZS on the ferroptosis‐induced mitochondrial damage has never been reported. This paper initially implied that XMZS can relieve neuronal damage after stroke by attenuating the ferroptosis‐induced mitochondrial damage and dysfunction. This discovery provided a reliable basis for the clinical application of XMZS in stroke treatment.

NRF2/GPX4/SLC7A11 pathway exerts a key effect on ferroptosis. NRF2 possesses an ability to activate the intracellular antioxidant defense system, then attenuating ferroptosis through up‐regulating GSH, GPX4, and SLC7A11 (Kang et al. 2025; Kose et al. ²⁰²²). GPX4 is a GSH‐regulated lipid repair enzyme, which can repress ferroptosis via reducing ROS and lipid peroxidation product accumulation (Zhou et al. 2023). SLC7A11 is capable of inducing the synthesis of GSH and then GPX4‐mediated detoxification of lipid peroxidation, thereby blocking ferroptosis (Koppula et al. 2021). The blocked activity of NRF2/GPX4/SLC7A11 pathway has been found in stroke, while its activation exerts neuroprotective effects and reduces cerebral infarct volume (Zhu et al. 2024). In this article, the blocked NRF2/GPX4/SLC7A11 pathway in MCAO rats and OGD/R‐induced primary hippocampal neurons was prominently activated by XMZS treatment. ML385 is an inhibitor of NRF2, which is capable of blocking the transcriptional activity of NRF2 (Shi et al. 2022). In this study, ML385 was utilized to treat MCAO rats and OGD/R‐induced primary hippocampal neurons. It was found that ML385 eliminated the inhibition of XMZS on ferroptosis in MCAO rats and OGD/R‐induced primary hippocampal neurons. Thus, XMZS exerted its efficacy in stroke by suppressing neuronal ferroptosis through activating the NRF2/GPX4/SLC7A11 pathway. The discovery had important implications for applying XMZS in treating stroke.

Taken together, this paper demonstrated the efficacy of XMZS in treating stroke. To elaborate, XMZS might attenuate stroke by suppressing neuronal ferroptosis via activating the NRF2/GPX4/SLC7A11 pathway. These findings provided evidence and a reliable molecular foundation for the clinical application of XMZS in stroke therapy. Future in‐depth animal studies on XMZS for stroke treatment and further exploration of its related molecular mechanisms will greatly advance the clinical application and promotion of XMZS in the field of stroke treatment.

This study has limitations. In in vitro study, XMZS medicated serum at different volume percentages was utilized to treat hippocampal neurons. Although the preparation of XMZS medicated serum maintained consistency in active ingredients as much as possible, measuring the concentration of active ingredients in XMZS medicated serum at different volume percentages would enhance the rigor of the research. However, this experiment cannot currently be conducted due to laboratory constraints. It will be a key focus of our future research.

5. Conclusion

This study implicated the therapeutic effect of the Traditional Chinese herbal formula XMZS on treating stroke. XMZS improved neurological function, diminished cerebral infarction volume, and mitigated neuronal damage of rats post‐stroke. From in vivo and in vitro mechanistic studies, XMZS was suggested to relieve stroke through mitigating neuronal ferroptosis by activating the NRF2/GPX4/SLC7A11 pathway. These findings proposed the Traditional Chinese herbal formula XMZS as an efficient way to manage stroke clinically.

Funding

This research was supported by the Natural Science Foundation of Hunan Province (Grant 024JJ9527); the Traditional Chinese Medicine Research Project of Hunan Province (Grant C2024029); the Natural Science Foundation of Changsha (Grant kq2502318); and the Scientific Research Project of the Changsha Municipal Health Commission (Grant KJ‐B2023080).

Supporting information

Figure S1: The base peak chromatograms of control serum, drug serum and drug in positive ion mode.

Figure S2: The base peak chromatograms of control serum, drug serum and drug in negative ion modes.

Figure S3: The serum active components of XMZS.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.